A bottoms-up method, within Dunford's Five-Part Positioning Framework, for defining an ideal customer profile (ICP): rather than guessing at firmographic targeting criteria up front, identify which existing or target accounts most value the distilled value themes produced by the Capability-to-Value "So What?" Exercise, then extract the shared traits of those accounts (industry, size, maturity, buying trigger) as the ICP definition.
The direction of inference matters — value themes first, matching accounts second, targeting criteria last — rather than starting from a demographic guess and hoping the value story fits.
Segway launched for a broad "getting around" market and mostly failed — an 80%-failure-rate outcome, in Olsen's words. But it wasn't a total flop: mall cops and tour groups adopted it. Working backward from who actually adopted, rather than who was pitched, reveals the real target segment and the reason they valued it — in this case, people who cover unusually high walking distances as part of their day.
A simplified proxy for segment membership: Olsen suggests a single behavioral question — "how many miles do you walk per day?" — as a rough stand-in for a fuller segmentation study, useful when a quick filter for target-market membership is needed rather than a full study.
See Two Competitive Strategies: Upper-Left Capture vs. "To-11" Disruption for why Segway's broader failure and narrow niche success both trace back to the same quadrant-position problem.